Intelligent State Estimation for Online Optimizing Control of a Reactor System exhibiting Input Multiplicity

Abstract

Combinations of real-time optimization (RTO) and model predictive control (MPC) have been widely employed in the process industry for tracking the economic optimum in the face of drifting disturbances and parameters. Online update of model parameters is a critical step in the implementation of RTO. In this work, an intelligent state and parameter estimation approach is developed by combining a fault diagnosis approach with a moving window-based online state and parameter estimator. The estimation of unmeasured disturbance(s)/ parameter(s)/ sensor bias(es) is carried out only when required and triggered by the fault identification scheme. Thus, the subset of parameters/faults that are being estimated online can change with time. This can avoid difficulties that arise due to the observability condition. The intelligent state and parameter estimator is further combined with an online optimizing control scheme consisting of integrated frequent RTO and adaptive MPC. The integrated scheme has embedded intelligence to auto-correct models used for estimation, control, and optimization and to decide whether the detected changes require the invocation of RTO. The efficacy of the proposed scheme is investigated using a benchmark CSTR system that exhibits input multiplicity behavior. The optimum operating point of this system is sensitive to mean shifts in unmeasured disturbances or system parameters. The proposed approach successfully isolates the parameter/ unmeasured disturbance/ sensor bias that has undergone abrupt change and tracks the shifting economic optimum without significant delays. Thus, the proposed integrated approach has the ability to handle normal, o¤-normal, and abnormal operating envelopes of the system.